CIMHA - Cluster Installer Monitor & Health Analyser
What is CIMHA :
CIMHA project is aimed to create an integrated suite to provision and
manage HA clusters in virtual environment. It will also constantly analyze
kernel/hardware health of cluster nodes by using
VESPER (a Linux kernel prober).
It will invoke (live) migration of Resources (like VMs) based on predictive
VESPER probe signals.
Project Summary :
CIMHA will provide a way to do the provisioning, monitoring and early detection
of failures of cluster nodes or resources, under a unified GUI.
ConVirt will
be used as the GUI infrastructure provider. CIMHA can provision physical as well as
Virtual Machines (VMs) remotely. It will be used to form and configure HA clusters
as well.
CIMHA will use VESPER to probe and monitor the health of the cluster nodes
and Virtual Machines (VMs). Based on this it will do the early detection of failures
and will act accordingly.
Features :
- OS installation
- VM creation, monitoring and (live) migration
- Cluster configuration and monitoring
- Analysing the Health of the nodes and VMs
Image :
Workflow (for Monitoring) :
- Cluster Manager in host requests Resource Manager for the status of services
in the VMs.
- Resource manager checks the status by communicating with proxy present
in VMs through XENBUS.
- The status of service is sent to the resource manager through XENBUS.
- Resource manager gathers the status information and sends it to the
Cluster manager.
- Cluster Manager in host requests for the health of host and VMs to
VESPER.
- VESPER communicates with SystemTap agent in VMs through XENBUS and
gathers information.
- VESPER sends back the information to Cluster manager.
- Based on the information from Resource Manager and VESPER, Cluster Manager
will do appropriate action.
Core Components :
GUI APIs :
- Web based APIs
- Python + GTK
Links :